In a wavelength division multiplexing (WDM) optical network currently used, there are plural paths in many cases until reaching a destination because the optical network has a mesh-like intricately complex configuration. Accordingly, plural candidates for a bypassing path (backup line) exist when a currently used line has failed. Thus, the WDM optical network currently used may have a good network configuration adaptable for, e.g., secondary and tertiary failures.
On the other hand, an optical fiber accompanies a factor that degrades a transmission signal, the factor being called “chromatic dispersion”. To avoid such degradation of the transmission signal, dispersion compensation is generally performed on wavelengths of signal light at each relay site with batch processing in, e.g., long-distance transmission over many spans. Examples of the relay site include an optical add-drop multiplexer (OADM) node and an in-line amplifier (ILA) node.
One of references regarding the related art is DAIKOKU Masahiro, “Polarization Mode Dispersion Suppression Technique for Super-High Speed Light Signal in Laid Optical Transmission Path”, February 2008.
Another related-art technique is disclosed in Japanese Laid-open Patent Publication No. 2008-010971.
However, when a signal of the type less endurable to the chromatic dispersion in 40G phase modulation is transmitted, for example, the chromatic dispersion is not sufficiently compensated for in some channel with the above-described batch compensation of the chromatic dispersion alone. Accordingly, the signal is further compensated individually for each wavelength depending on an amount of a total chromatic dispersion after having passed through a network (hereinafter referred to as “residual chromatic dispersion”). Thus, the signal is appropriately compensated such that the chromatic dispersion is held within a tolerance range enabling the signal to be received by a receiver. When a failure occurs in the network and the path is changed over to a bypassing path, the residual chromatic dispersion is changed and individual degrees of chromatic dispersion compensations are also changed. This raises the problem that the transmitted signal is not received by the receiver and an error state continues until the individual compensations are optimized.
Meanwhile, if a value of the residual chromatic dispersion in the bypassing path, which is to be taken in the occurrence of a failure, is known in advance, the individual degrees of chromatic dispersion compensations can be properly set as soon as the communication path is changed to the bypassing path. In more detail, when there is one bypassing path, for example, the individual extent of chromatic dispersion compensation can be properly set in a moment by preparing respective residual chromatic dispersion values for the currently used path and the bypassing path. In a mesh network, however, because the bypassing path is present in plural or infinite number, it is not practical to prepare respective residual chromatic dispersion values for the plural paths that may be used.